High-speed and Cost Effec tive Machine Vision System within the Industry of Preserved Vegetabl es C. Fernandez, P. J. Navarro, J. Suardiaz, M. Jimenez, A. Iborra. UNIVERSIDAD POLITECNICA DE CARTAGENA DSIE CAMPUS MURALLA DELMAR, S/N 30201CARTAGENA, SPAIN {carlos.fernandez, juan.suardiaz, pedroj .navarro, manuel.j imenez, andres.iborra}@upct.es Abstrac t The paper describes a sophisticated and low cost Automated Visual Inspection System (AVIS) for qualitycontrol of preserved orange segments, widely applicable to production processes of preserved fruits and vegetables. Main constraints concerning these k ind of inspection applications are addressed: the need of on- line operation together withan strong requirement of economic profitability . The strong commitment of above circumstances have forced the development of a flexible and low cost AVIS architecture. An special effort has been put in the design of the defect detection algorithms to reach two main obj ectives: accurate feature extraction and on- line capabilities, bothconsideringrobustness and low processing time. An on- line implementation has been possible byusingIP cores over re- configurable FPGA based BallyNue card, capable to inspect up to ten orange segments by second. 1 Introduc tion Along the last years more and more applications arise for visual inspection of many different products with on-line achievement [Jain, 1995][Jyrkinen, 1999][Kj aergaard, 1999] . This has been possible mainly due to the arrival of more powerful tools (hardware and software) into the vision developers community at better prices. These applications have not yet massively reached -although partial solutions have been reported, like fruit sorting- the fruit/vegetable processing industry due to several reasons: complexity of the inspection task, traditional idiosyncrasy within the manufacturers and market particularities for this kind of production. Production workshops are highly automated ( raw fruit classification, washing, peeling, separation into different sizes,packaging…) being the inspection task the only one that is still performed by skilled operators,which represents the 30% of the man power in a typical workshop. Inspection mistakes and variability,cost to the Spanish producers of preserved fruit/vegetables 10Meuro per year ( aprox . 5% of 50 .000Tm produced every year). Standards as Quality Standards for Preserved Vegetables Exportation or Quality Standards for Preserved Vegetables, among others, set commercial categories for preserved orange segments according to maximum percentage of broken slices and peeling defects. The challenge is to improve present manual classification by means of an automated system which assures on-line adaptation to production conditions. 2 Spec ifications forthe visual inspec tion task The process of visual inspection at each workshop where preserved vegetables/fruits are prepared is made at the final stage,after the vegetables/fruits have been peeled and separated into pieces. In our case study –preserved orange segments- this inspection is achieved by four to sixexpert operators placed along both sides of a conveyor belt which carries the segments with a typical speed of 2m/min with a global inspection rate of about ten segments by second. Illustration 1 show images from orange segments with different quality. There are different product qualities;final quality corresponds to a maximum content of broken segments. No strange elements are permitted ( such as peels, leaves, ...), neither are the so called double segments. a) Good segments b) Broken segments MVA2005 IAPR Conference on Machine VIsion Applications, May 16-18, 2005 Tsukuba Science City, Japan 9-2 372
4
Embed
High-speed and Cost Effective Machine Vision System within the …b2.cvl.iis.u-tokyo.ac.jp/mva/proceedings/Commemorative... · 2008-02-25 · b) inner evolving c) extern evolving
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
High-speed and Cost Effective M achine Vision System within the
Industry of Preserved Vegetables
C. Fernandez, P. J. Navarro, J. Suardiaz, M . Jimenez, A. Iborra.